Upload
phamtram
View
214
Download
0
Embed Size (px)
Citation preview
IC1004 / Newcom# School:
"Beyond 4G Networks in Cities: From Theory to Experimentation and Back”
November 25 -‐ 28, 2013 CTTC, Castelldefels, Spain
Abstracts of Lectures and Description of Lab Sessions
Werner Mohr (NSN), T1: Research Challenges for B4G Mobile and Wireless Communications Systems
Mobile and wireless communication are growing further globally and will offer a plethora of services to users. Broadband mobile communication systems like LTE are currently being deployed. Communication networks are also increasingly be used for vertical sectors to manage utility services like energy, gas and water systems and are critical infrastructures for our societies and economies. In particular these applications are based on sensor-‐, IoT-‐ or M2M-‐based systems with a huge number of connected devices, which will be part of future networks. These developments are resulting in exponential traffic growth in the coming years. Future networks will increasingly be based on heterogeneous radio access systems from short range to wide range communications, from low to broadband throughput rates and from low to higher possible latencies in order to support all envisaged application scenarios.
The expected traffic growth and the huge number of connected devices results in challenging requirements on the energy efficiency of future systems in order to keep the energy and CO2 footprint of future systems as low as possible. The availability of sufficient frequency spectrum is a major concern to support the necessary system capacity. Additional challenges are significantly reduced latency for improved user experience, heterogeneous systems, self-‐organizing networks to cope with the system complexity, a more flexible use of the frequency spectrum and cloud-‐based architectures. Cognitive networks are complementing future systems in order to improve reliability, quality and to reduce operational and energy cost. This is creating huge challenges for research, design and deployment of future communication networks. The network architecture will have to undergo significant changes to meet requirements also to support economic service provision. It will become more flat, virtualization of network resources and cloud concepts reduce capital (CAPEX) investment and operational expenditures (OPEX) and will allow faster and easier introduction of new services and thereby innovation in service offering. Such Software-‐Defined Networking (SDN) concepts in combination with cognitive systems will support self-‐organized systems in order to manage the overall network complexity.
Future communication networks require a holistic approach in research, standardization, regulations and system development. Research initiatives have started in different regions of the world. In Europe the new Framework Program Horizon 2020 will address this area in coordinated research activities and in particular in a 5G Public Private Partnership on communication network infrastructure.
The presentation will present the trends towards future systems, the expected research challenges and the upcoming 5G Public-‐Private-‐Partnership.
Fabio Dovis (PoliTO), T2: Localisation by Satellite Systems The class will provide the basic principles of Satellite Based localisation, describing the signals and the main stages of the receivers. The presentation will address also the main factors limiting the performance and it will discuss the opportunities for the hybridisation with terrestrial communication networks to extend a reliable positioning service to harsh environments.
Chiara Buratti (UniBO), T3: On the Impact of Medium Access Control Protocols on Multi-‐Hop Networks
The lecture will be split into two parts: the first part will provide some basics on Medium Access Control (MAC) and routing protocols for wireless sensor networks (WSNs), and a second part which will be focused on the IEEE 802.15.4 and Zigbee standard and on smart city applications.
In particular, in the first part some of the most suitable contention-‐based MAC protocols (e.g., CSMA/CA) and routing protocols (e.g., AODV), for WSNs will be described in details. The second part of the lecture will focus on protocols for smart city applications, having the IEEE 802.15.4 and Zigbee standard as reference. Some case studies, providing some real-‐world implementations, will be described and numerical results achieved through experimentation on the field will be shown. The latter will provide some insight on the interplay between theory and experimentations and the impact of MAC on the routing protocols and on the topologies generated in the networks.
Reiner Thomae (Ilmenau), T4: Experimental Propagation Studies Using Wideband Real-‐Time MIMO Channel Sounding
Multidimensional sounder system architecture, choice of transmit signals, estimation of delay Doppler response, sounding antennas and antenna arrays, application of sounding data, high resolution parameter estimation, multidimensional data model, iterative ML parameter estimation, antenna data model.
Nicola Baldo (CTTC), T1: Simulation of LTE Systems in ns-‐3
This lecture is aimed at researchers who are interested in the simulation of LTE systems for their research activities, and who have no prior experience with ns-‐3. The lecture will first provide a generic overview of the ns-‐3 simulator, discussing its architecture, its main features, and its models library. After this introduction, the LTE module of ns-‐3 will be presented, discussing its modeling assumptions first, and then examining its components (radio protocol stack, PHY model, MAC Scheduler API & models, EPC model, buildings & propagation models).
Nikolaos Bartzoudis (CTTC), T1 Seminar: "Prototyping the Physical Layer of Modern Wireless Communication Systems: Development Flows, Challenges and Pitfalls"
The processing complexity of a PHY-‐layer prototype may vary depending on the target scenario that needs to be developed and experimentally validated. A critical factor towards this end is if the prototype should feature offline, semi real-‐time or full real-‐time operation. Other parameters that play an important role in the low-‐level design and implementation of the prototype include the signal bandwidth, the use of multi-‐antenna schemes, the receiver mobility and the degree (or rate) that a signal needs to be re-‐formatted/re-‐constructed in order to yield adaptive/flexible communications. The influence of the mentioned factors and parameters needs to be thoroughly analyzed in order to choose the most adequate prototyping methodology which could be either based on a software (e.g. SDR), hardware-‐software, or hardware (e.g., FPGA or ASIC design) approach. Moreover, the specifications, operating conditions (including assumptions and simplifications) and configuration of the selected experimental setup greatly affect the validity of the conducted benchmarking for each prototype. Apart from the low-‐level design, implementation and performance objectives the researcher/engineer/student needs to deal with other heterogeneous challenges such as the learning curve of the selected methodology, the development time/cost and the modularity/reusability/scalability/portability of the prototype. Additionally, industry trends (market manipulation) or popular academic initiatives usually impair the correct selection of a suitable methodology tailored for the specific needs of a research and development project. Thus, the goal of this seminar is to give an overview of the different approaches used across the academic community to prototype the Physical Layer of wireless communication systems trying to highlight the associated challenges and limitations that apply in each case. Likewise it is intended to convey the message that each PHY-‐layer prototype requires a certain development flow and thus a specific research/engineering modus operandi.
Jesus Alonso-‐Zarate (CTTC), T1 Seminar: "Machine-‐to-‐Machine Technologies and Smart Cities"
The unprecedented communication paradigm of machine-‐to-‐machine (M2M), facilitating 24/7 ultra-‐reliable connectivity between a prior unseen number of automated devices, is currently gripping both industrial as well as academic communities. The aim of this tutorial is to provide an academic, technical and industrial insight into latest key aspects of wireless M2M networks, with particular application to smart cities and smart grids. We will provide an in-‐depth introduction to the particularities of M2M systems, and then dwell in great depths on the capillary and cellular embodiments of M2M. The focus of capillary M2M will be on IEEE (.15.4e) and IETF (6LoWPAN, ROLL, COAP) standards compliant low-‐power multihop networking designs; furthermore, for the first time, low power Wifi will be dealt with and positioned in the eco-‐system of capillary M2M. The focus of cellular M2M will be on latest activities, status and trends in leading M2M standardization
bodies with technical focus on ETSI M2M and 3GPP LTE-‐M; furthermore, we will discuss analytical and simulation works quantifying the performance and impact of M2M in legacy cellular networks.
Along the entire tutorial, challenges and open issues will be identified, thus making the material presented in this tutorial useful for industry and inspiring for researchers and academics alike.
Claude Oestges (UCL), T4: Reference and Standardized Channel Models
This lecture will propose an overview of recent standardized and reference models for beyond 4G networks in urban environments. In particular, the lecture will highlight how most models essentially rely on a combination of experimentation and smart data processing. Covering and comparing the recent modeling efforts in WINNER and COST 2100, the lecture will also point out at the various modeling challenges.
Florian Kaltenberger (Eurecom), T4: Physical Layer Abstraction for LTE
Physical layer abstraction is the process of modeling the performance of the physical layer (in terms of block error rates or throughput) as a function of the radio channel without running the time consuming MODEM and the channel convolution. Such models are useful for two different purposes: Firstly they can be used in the implementation of a user equipment (UE) to compute the feedback (channel quality information 0 CQI) and secondly they can be used in large-‐scale system level simulations to speed up simulation time.
In this 1.5h course we will study the basics of physical layer abstraction with the application example of turbo codes in LTE. We will cover both Exponential effective SINR mapping (EESM) and Mutual Information based SINR mapping (MIESM). We will describe the calibration process in detail and give some practical examples using the OpenAirInterface simulation tools. We will conclude the course by applying the learned techniques to the LTE feedback calculation and fast link adaptation and system level simulations.
Ronald Raulefs (DLR), T2: Network Based Localisation
Recent years have seen an explosion of location based services. On the other hand, the limitations of GPS for indoor, urban canyons etc. have led to an evolution of existing networks (e.g. LTE) to provide network based location estimation. Today’s and future mobile radio devices provide a heterogeneous portfolio of radio access technologies. We provide an overview of recent advances in cooperative communication based location estimation, fingerprinting techniques that also work in non line-‐of-‐sight, the exploitation of inertial modalities and self-‐learning techniques to enable context-‐aware network adaptions based on inferred radio maps.
Roberto Verdone (UniBO), T3: B4G Networks and the Smart 2020 City
Smart City applications need to rely on flexible and high density radio access networks deployed in the urban context. Proper integration of short range, B4G cellular and multi-‐hop wireless techniques, is required. The first part of the lecture will discuss the technical and economic/practical aspects of the deployment of radio network infrastructures for Smart City applications. In the second part, an interactive/brainstorming session will be dedicated to the definition of the possible technological options and the open research problems that need to be addressed within Horizon2020. The lecture will serve somehow as a recap of the various topics addressed during the school.
Carles Fernandez (CTTC), Javier Arribas (CTTC) , T2: Hybrid localisation: from satellite to heterogeneous localisation techniques for Beyond 4G Networks
Attendees to Track T2 will be exposed to a hands-‐on approach to the signal processing involved in satellite-‐based navigation receivers, working with real-‐life signals and replacing the traditional, 'black-‐box' integrated circuit receiver by GNSS-‐SDR, an open source project that implements a global navigation satellite system software defined receiver in C++. With GNSS-‐SDR, users can build a GNSS software receiver by creating a graph where the nodes are signal processing blocks and the lines represent the data flow between them. The software provides an interface to different suitable RF front-‐ends and implements the entire receiver’s chain up to the navigation solution. Its design allows any kind of customization, including interchangeability of signal sources, signal processing algorithms, interoperability with other systems, output formats, and offers interfaces to all the intermediate signals, parameters and variables. This allows a scrutinization of all the stages
involved in GNSS signal processing, provides a clear view of how a GNSS receiver works internally and constitutes a platform for further research and experimentation.
Attendees are supposed to be familiar with basic digital signal processing concepts and techniques, as well as having some previous exposure to C++ and software design patterns. Programming will not be required, but familiarity with the language could make the course more profitable. A background in localisation systems is desirable but not a requirement, since all the key aspects will be explained in detail before the experimental activity.
Chiara Buratti (UniBo), T3: IoT component in the 2020 City: From MAC to Routing Aspects
During the laboratory experience students will use the EuWIn facilities and in particular the FLEXTOP platform, composed of 100 IEEE 802.15.4 / Zigbee devices deployed in a corridor at the University of Bologna. The platform, in fact, is remotely accessibly through a VPN connection. Students will have the opportunity to modify a software implementing a smart city application, download the modified software on the devices in the FLEXTOP platform and wait for the results, to process them.
The throughput in a multi-‐hop point-‐to-‐point network will be first computed through experimentations. While as a second experiment, a smart city application will run on 28 devices. Results in terms of topologies generated in the network and packet loss rate will be analyzed and processed.The first experiment will mainly aim at demonstrating the importance of the interplay between theory and experimentations, while the second experiment will be much more focused on the interplay between MAC and routing.
UniBo will provide a virtual machine to be installed on the PCs for the LAB session.The virtual machine could be based on Oracle Virtual Box (preferred) or VMware VM player depends on your requirements (see below). The PCs were the virtual machines (VM) will be installed (for a maximum number of 5, assuming to have 5 groups composed of 3 people each), should be able to access to Internet, in particular a VPN session will be started on each VM. One possible solution is that CTTC provides 5 PCs with the VM environment installed and the Internet connection. Alternatively, students can use their own laptops, however CTTC should provide to students a proper Internet connection which ensures to establish a VPN session.
Christian Schneider (TU Ilmenau), T4: Derivation of Large Scale Parameter for WINNER channel models based on Data Sets from the Ilmenau Reference Scenario (Urban Macro Cell)
So called large scale parameters (LSPs) are the scenario dependent control parameters for stochastic geometry based channel models as from the SCM or WINNER family. To provide the models with reliable statistical distributions of these parameters is important to derive them based on channel sounding data sets from appropriate scenarios. Within the lab session the theoretical background as well as measurement data related aspects will be highlighted. The experimental part will cover the implementation and derivation of the required power und delay domain parameters as path loss, shadow fading, K-‐factor, delay spread and XPR as well as their de-‐correlation und cross-‐correlation statistics.